Model Anatomy See what's inside open-weight AI models

Can I run Qwen3.8-27B on an RTX 6000 Ada?

Yes, with a GGUF version

The published weights are too big, but the Q8_0 file from bartowski/Qwen3.8-27B-GGUF (29.1 GB) fits, with room for ≈ 208K tokens of context.

Full analysis of Qwen3.8-27B Everything that runs on RTX 6000 Ada Get the Q8_0 file on Hugging Face
ollama run hf.co/bartowski/Qwen3.8-27B-GGUF:Q8_0

Files and context on RTX 6000 Ada

Each file of Qwen3.8-27B (and of GGUF versions the site has analyzed), and how much context fits next to it, with the context cache in 16-bit or 8-bit.

FileSizeOn RTX 6000 Ada Context (16-bit)Context (8-bit)
As published55.6 GBDoesn't fit——
BF16ggml-org/Qwen3.8-27B-GGUF53.8 GBDoesn't fit——
Q8_0ggml-org/Qwen3.8-27B-GGUF28.6 GBFits215K256K
Q4_K_Mggml-org/Qwen3.8-27B-GGUF19.4 GBFits256K256K
BF16bartowski/Qwen3.8-27B-GGUF54.7 GBDoesn't fit——
Q8_0bartowski/Qwen3.8-27B-GGUF29.1 GBFits208K256K
Q6_K_Lbartowski/Qwen3.8-27B-GGUF25.0 GBFits256K256K
Q6_Kbartowski/Qwen3.8-27B-GGUF23.9 GBFits256K256K
Q6_K_Sbartowski/Qwen3.8-27B-GGUF22.9 GBFits256K256K
Q5_K_Lbartowski/Qwen3.8-27B-GGUF21.5 GBFits256K256K
Q5_K_Mbartowski/Qwen3.8-27B-GGUF20.9 GBFits256K256K
Q5_K_Sbartowski/Qwen3.8-27B-GGUF19.6 GBFits256K256K
Q4_K_Lbartowski/Qwen3.8-27B-GGUF18.8 GBFits256K256K
Q4_1bartowski/Qwen3.8-27B-GGUF17.8 GBFits256K256K
IQ4_NLbartowski/Qwen3.8-27B-GGUF17.4 GBFits256K256K
Q4_K_Mbartowski/Qwen3.8-27B-GGUF17.4 GBFits256K256K
Q3_K_XLbartowski/Qwen3.8-27B-GGUF16.4 GBFits256K256K
Q4_K_Sbartowski/Qwen3.8-27B-GGUF16.4 GBFits256K256K
Q4_0bartowski/Qwen3.8-27B-GGUF16.3 GBFits256K256K
IQ4_XSbartowski/Qwen3.8-27B-GGUF15.5 GBFits256K256K
IQ3_Mbartowski/Qwen3.8-27B-GGUF14.9 GBFits256K256K
Q3_K_Lbartowski/Qwen3.8-27B-GGUF14.1 GBFits256K256K
Q3_K_Mbartowski/Qwen3.8-27B-GGUF13.4 GBFits256K256K
Q2_K_Lbartowski/Qwen3.8-27B-GGUF13.1 GBFits256K256K
IQ3_XSbartowski/Qwen3.8-27B-GGUF12.8 GBFits256K256K
Q3_K_Sbartowski/Qwen3.8-27B-GGUF12.7 GBFits256K256K
IQ3_XXSbartowski/Qwen3.8-27B-GGUF12.3 GBFits256K256K
Q2_Kbartowski/Qwen3.8-27B-GGUF10.8 GBFits256K256K
IQ2_Mbartowski/Qwen3.8-27B-GGUF10.5 GBFits256K256K
IQ2_Sbartowski/Qwen3.8-27B-GGUF9.68 GBFits256K256K
IQ2_XSbartowski/Qwen3.8-27B-GGUF9.09 GBFits256K256K
IQ2_XXSbartowski/Qwen3.8-27B-GGUF8.88 GBFits256K256K
IQ4_XSbyteshape/Qwen3.8-27B-GGUF13.1 GBFits256K256K
IQ3_Sbyteshape/Qwen3.8-27B-GGUF11.0 GBFits256K256K
IQ3_XSbyteshape/Qwen3.8-27B-GGUF10.4 GBFits256K256K
IQ3_XXSbyteshape/Qwen3.8-27B-GGUF9.92 GBFits256K256K
IQ2_XXSbyteshape/Qwen3.8-27B-GGUF8.84 GBFits256K256K

Qwen3.8-27B on other hardware

Popular models that run on RTX 6000 Ada

Assumes 90% of the card's memory is usable and at least 4K tokens of context for one conversation, with the context cache in 16-bit unless noted. Fits leaves at least 20% of that memory for context; Tight leaves less.